Robust coexistence in simple microbial communities

This repository contains data, code, and analysis notebooks for a study of robust coexistence in simple microbial communities. The work examines how four bacteria assemble and coexist across varying carbon source environments, and uses a consumer-resource model (CRM) to explore the mechanistic basis of coexistence.

Species

Abbreviation	Species
At	Agrobacterium tumefaciens
Ct	Comamonas testosteroni
Ml	Microbacterium liquefaciens
Oa	Ochrobactrum anthropi

Repository structure

data/                   Experimental data, organized by experiment
notebooks/              Jupyter notebooks for analysis and figure generation
python/                 Python scripts for consumer-resource model (CRM) simulations
simulation_results/     Output files from CRM parameter sweeps and dFBA simulations

Reproducing the analysis

The numbered folders in data/, notebooks/, and Figures/ follow the same ordering:

1. Growth phenotyping of monocultures
2. First community assembly experiment
3. Second community assembly experiment
4. Chemostat coexistence experiments
5. Pairwise interaction measurements
6. Consumer-resource model (CRM) parameterization and simulation

Run notebooks in order within each section. CRM parameter sweeps (section 6) are run via scripts in python/6_CRM/ and results are saved to simulation_results/.

Dependencies

• Python 3.x
• numpy, scipy, pandas, matplotlib, seaborn
• joblib (for parallel CRM sweeps)
• Jupyter for notebooks

Cite our work

Snorre Sulheim, Miguel Teixeira, Eric Ulrich, Alisson Gillon, Samuel Testa, Prajwal Padmanabha, Daniel Machado & Sara Mitri (2026). Cross-feeding enables robust coexistence between four bacterial species. bioRxiv 2026.04.20.719729